Microsoft AI-300 Practice Test 76
Question 750
While preparing a secure Foundry model or agent release, the following condition occurs: A nightly GenAIOps evaluation workflow stops connecting to its Foundry project after the release team revokes an old client secret used by the CI job. The remediation must meet enterprise access and reliability requirements without a broad bypass. What approved correction resolves the failure without modifying unrelated healthy services?
- Define the model and project deployment lifecycle dependencies and verify provisioning completion before dependent app configuration.
- Grant the narrowly scoped search data access role to the correct agent identity and use identity-based connection configuration.
- Migrate the workflow to supported managed or federated identity, or rotate the credential through approved secret management and verify minimal access.
- Select an approved multimodal model or add a supported image-to-text preprocessing service before the text generation step.
Explanation
Correct Answer
C:Migrate the workflow to supported managed or federated identity, or rotate the credential through approved secret management and verify minimal access. This targets foundry service principal credential rotated without updating pipeline. CI authentication is a separate dependency of a GenAIOps deployment; token renewal and secret rotation require an explicit supported credential flow. Verify the failure through: Inspect workflow identity claims, login failure logs, secret references, federated identity options and target resource permissions. Validate success with: The workflow authenticates through a governed identity after old secrets are disabled and completes the intended evaluation without privileged keys in source control. This matters here because while preparing a secure Foundry model or agent release, the following condition occurs: A nightly GenAIOps….
Incorrect Answers
A:Define the model and project deployment lifecycle dependencies and verify provisioning completion before dependent app configuration. This could address foundry bicep rollout omits model deployment dependencies, but it does not solve the specified foundry service principal credential rotated without updating pipeline failure. The matching remediation is: Migrate the workflow to supported managed or federated identity, or rotate the credential through approved secret management and verify minimal access.
B:Grant the narrowly scoped search data access role to the correct agent identity and use identity-based connection configuration. This could address agent runtime managed identity cannot read retrieval data, but it does not solve the specified foundry service principal credential rotated without updating pipeline failure. The matching remediation is: Migrate the workflow to supported managed or federated identity, or rotate the credential through approved secret management and verify minimal access.
D:Select an approved multimodal model or add a supported image-to-text preprocessing service before the text generation step. This could address vision input sent to text-only foundation model, but it does not solve the specified foundry service principal credential rotated without updating pipeline failure. The matching remediation is: Migrate the workflow to supported managed or federated identity, or rotate the credential through approved secret management and verify minimal access.
Question 751
While optimizing a production RAG or fine-tuned model service, the following condition occurs: A customer support model fine-tuned on historical chats becomes inconsistent because examples mix strict JSON tool calls with casual prose for identical task types. The resulting model must meet held-out quality and security criteria. Select TWO actions: one decisive diagnostic and its corresponding targeted remediation.
- Retain governed immutable fine-tune artifacts and record deployment lineage with tested restore/rollback procedures.
- Filter or downrank superseded records before hybrid retrieval and use governed version metadata for source selection.
- Clean and stratify the supervised fine-tuning data to a coherent supported task format before retraining and evaluating.
- Inspect fine-tuning records, role/message formatting, target response schemas, label consistency and curated holdout behavior.
- Inspect information update frequency, retrieval freshness requirement, model task errors, training cost and data-change workflow.
Explanation
Correct Answers
C:Clean and stratify the supervised fine-tuning data to a coherent supported task format before retraining and evaluating. This targets supervised fine-tuning dataset includes conflicting assistant behaviors. Supervised fine-tuning learns patterns in labeled examples; contradictory targets can degrade reliability even when many examples are provided. Verify the failure through: Inspect fine-tuning records, role/message formatting, target response schemas, label consistency and curated holdout behavior. Validate success with: The customized model returns the approved structured behavior on held-out scenarios without a spike in schema violations.
D:Inspect fine-tuning records, role/message formatting, target response schemas, label consistency and curated holdout behavior. This is the most relevant evidence for supervised fine-tuning dataset includes conflicting assistant behaviors. Supervised fine-tuning learns patterns in labeled examples; contradictory targets can degrade reliability even when many examples are provided. The matching controlled remediation is: Clean and stratify the supervised fine-tuning data to a coherent supported task format before retraining and evaluating.
Incorrect Answers
A:Retain governed immutable fine-tune artifacts and record deployment lineage with tested restore/rollback procedures. This is associated with fine-tune production endpoint references deleted training artifact, not the essential evidence-and-fix pair for supervised fine-tuning dataset includes conflicting assistant behaviors. First examine: Inspect fine-tuning records, role/message formatting, target response schemas, label consistency and curated holdout behavior. Then correct: Clean and stratify the supervised fine-tuning data to a coherent supported task format before retraining and evaluating.
B:Filter or downrank superseded records before hybrid retrieval and use governed version metadata for source selection. This is associated with hybrid retrieval double-counts outdated documents, not the essential evidence-and-fix pair for supervised fine-tuning dataset includes conflicting assistant behaviors. First examine: Inspect fine-tuning records, role/message formatting, target response schemas, label consistency and curated holdout behavior. Then correct: Clean and stratify the supervised fine-tuning data to a coherent supported task format before retraining and evaluating.
E:Inspect information update frequency, retrieval freshness requirement, model task errors, training cost and data-change workflow. This is associated with fine-tuning task selected when retrieval would better handle changing facts, not the essential evidence-and-fix pair for supervised fine-tuning dataset includes conflicting assistant behaviors. First examine: Inspect fine-tuning records, role/message formatting, target response schemas, label consistency and curated holdout behavior. Then correct: Clean and stratify the supervised fine-tuning data to a coherent supported task format before retraining and evaluating.
Question 752
As an MLOps engineer reviews a model training or inference pipeline, the following condition occurs: An Azure ML production model serves predictions, but the operator reports daily accuracy without any actual outcome labels arriving until several weeks later. The resolution must preserve quality gates, documented artifacts and rollback options. Select TWO actions: one decisive diagnostic and its corresponding targeted remediation.
- Inspect prediction and ground-truth join keys, label availability delay, inference logging and published monitoring metrics.
- Monitor available proxy signals immediately and compute labeled performance only after verified outcomes arrive through a governed join.
- Configure a suitable seasonal baseline or context-aware drift monitor while retaining meaningful anomaly thresholds.
- Restrict access, redact or aggregate sensitive examples and publish a privacy-reviewed evaluation artifact for authorized reviewers.
- Inspect model retraining pipeline graph, evaluation outputs, baseline comparison, deployment step conditions and release records.
Explanation
Correct Answers
A:Inspect prediction and ground-truth join keys, label availability delay, inference logging and published monitoring metrics. This is the most relevant evidence for accuracy monitoring lacks delayed outcome labels. Model performance against ground truth cannot be measured correctly before outcomes are known; operational health signals are not substitutes for labeled accuracy. The matching controlled remediation is: Monitor available proxy signals immediately and compute labeled performance only after verified outcomes arrive through a governed join. This matters here because as an MLOps engineer reviews a model training or inference pipeline, the following condition occurs: An Azure ML….
B:Monitor available proxy signals immediately and compute labeled performance only after verified outcomes arrive through a governed join. This targets accuracy monitoring lacks delayed outcome labels. Model performance against ground truth cannot be measured correctly before outcomes are known; operational health signals are not substitutes for labeled accuracy. Verify the failure through: Inspect prediction and ground-truth join keys, label availability delay, inference logging and published monitoring metrics. Validate success with: Reports distinguish provisional drift/latency signals from later evidence-based accuracy and recall. This matters here because as an MLOps engineer reviews a model training or inference pipeline, the following condition occurs: An Azure ML….
Incorrect Answers
C:Configure a suitable seasonal baseline or context-aware drift monitor while retaining meaningful anomaly thresholds. This is associated with data drift monitor uses wrong reference season, not the essential evidence-and-fix pair for accuracy monitoring lacks delayed outcome labels. First examine: Inspect prediction and ground-truth join keys, label availability delay, inference logging and published monitoring metrics. Then correct: Monitor available proxy signals immediately and compute labeled performance only after verified outcomes arrive through a governed join. This matters here because as an MLOps engineer reviews a model training or inference pipeline, the following condition occurs: An Azure ML….
D:Restrict access, redact or aggregate sensitive examples and publish a privacy-reviewed evaluation artifact for authorized reviewers. This is associated with model evaluation exposes sensitive examples in explainability report, not the essential evidence-and-fix pair for accuracy monitoring lacks delayed outcome labels. First examine: Inspect prediction and ground-truth join keys, label availability delay, inference logging and published monitoring metrics. Then correct: Monitor available proxy signals immediately and compute labeled performance only after verified outcomes arrive through a governed join.
E:Inspect model retraining pipeline graph, evaluation outputs, baseline comparison, deployment step conditions and release records. This is associated with retraining pipeline deploys without model quality gate, not the essential evidence-and-fix pair for accuracy monitoring lacks delayed outcome labels. First examine: Inspect prediction and ground-truth join keys, label availability delay, inference logging and published monitoring metrics. Then correct: Monitor available proxy signals immediately and compute labeled performance only after verified outcomes arrive through a governed join.
Question 753
During a Foundry generative AI quality and observability incident, the following condition occurs: A support bot repeats its complete policy corpus in every prompt, increasing token costs and latency, although only a few retrieved paragraphs are relevant to each answer. The fix must preserve evaluation validity and avoid unnecessary user-data collection. Which targeted Azure AI operations change best addresses the documented problem?
- Instrument cost attribution across generation and embedding services, cache reusable embeddings and alert on unplanned spend.
- Bind relevance evaluation to the authentic user request and assess query rewriting as a separate retrieval-stage metric.
- Use relevant retrieval context, justified response limits and supported caching after verifying quality on evaluation examples.
- Mask sensitive tokens and personal data, apply least-privilege log access and retain only necessary structured debugging evidence.
Explanation
Correct Answer
C:Use relevant retrieval context, justified response limits and supported caching after verifying quality on evaluation examples. This targets prompt truncation strategy increases generated-token spend. Context volume affects token cost and latency; sending unnecessary text on every request is not equivalent to more reliable grounding. Verify the failure through: Inspect prompt token composition, repeated static sections, retrieval result counts, cache opportunities and groundedness metrics. Validate success with: Average tokens per successful request falls while groundedness and task success remain within accepted limits.
Incorrect Answers
A:Instrument cost attribution across generation and embedding services, cache reusable embeddings and alert on unplanned spend. This could address token cost report omits embedding requests, but it does not solve the specified prompt truncation strategy increases generated-token spend failure. The matching remediation is: Use relevant retrieval context, justified response limits and supported caching after verifying quality on evaluation examples.
B:Bind relevance evaluation to the authentic user request and assess query rewriting as a separate retrieval-stage metric. This could address relevance evaluator uses wrong intended user query, but it does not solve the specified prompt truncation strategy increases generated-token spend failure. The matching remediation is: Use relevant retrieval context, justified response limits and supported caching after verifying quality on evaluation examples.
D:Mask sensitive tokens and personal data, apply least-privilege log access and retain only necessary structured debugging evidence. This could address debug traces leak secrets and retrieved personal records, but it does not solve the specified prompt truncation strategy increases generated-token spend failure. The matching remediation is: Use relevant retrieval context, justified response limits and supported caching after verifying quality on evaluation examples.
Question 754
During a retrieval or model-customization performance review, the following condition occurs: A RAG search index blends keyword and vector results but weights retired versions of a procedure equally with current publications, so old policy text appears first. The change must keep citations, privacy and previously approved behavior intact. What should the engineer examine first to confirm the specific AI-300 workload issue?
- Check document effective dates, index filters, scoring profiles, vector/keyword rankings and available canonical document metadata.
- Inspect information update frequency, retrieval freshness requirement, model task errors, training cost and data-change workflow.
- Inspect representative code-based queries, BM25/keyword results, vector matches, analyzer configuration and fusion ranking behavior.
- Inspect top-k search scores, threshold, query normalization, relevant passages and test question recall.
Explanation
Correct Answer
A:Check document effective dates, index filters, scoring profiles, vector/keyword rankings and available canonical document metadata. This is the most relevant evidence for hybrid retrieval double-counts outdated documents. Hybrid scoring does not inherently enforce document freshness or authoritative version; metadata filtering is a separate ranking control. The matching controlled remediation is: Filter or downrank superseded records before hybrid retrieval and use governed version metadata for source selection. This matters here because a RAG search index blends keyword and vector results but weights retired versions of a procedure equally with….
Incorrect Answers
B:Inspect information update frequency, retrieval freshness requirement, model task errors, training cost and data-change workflow. This would investigate fine-tuning task selected when retrieval would better handle changing facts rather than the documented hybrid retrieval double-counts outdated documents issue. The relevant investigation is: Check document effective dates, index filters, scoring profiles, vector/keyword rankings and available canonical document metadata.
C:Inspect representative code-based queries, BM25/keyword results, vector matches, analyzer configuration and fusion ranking behavior. This would investigate pure vector search misses exact regulatory clause codes rather than the documented hybrid retrieval double-counts outdated documents issue. The relevant investigation is: Check document effective dates, index filters, scoring profiles, vector/keyword rankings and available canonical document metadata.
D:Inspect top-k search scores, threshold, query normalization, relevant passages and test question recall. This would investigate rag similarity threshold filters all relevant evidence rather than the documented hybrid retrieval double-counts outdated documents issue. The relevant investigation is: Check document effective dates, index filters, scoring profiles, vector/keyword rankings and available canonical document metadata. This matters here because a RAG search index blends keyword and vector results but weights retired versions of a procedure equally with….
Question 755
During a Microsoft Foundry generative AI deployment review, the following condition occurs: A document-assistant project exists in Foundry but cannot infer because its configured model deployment name is from a different project and unavailable to the target identity. The deployment must retain governed identities, endpoint contracts and auditable release history. Which investigation best verifies the cause before changing production configuration?
- Compare experiment assignments, model deployment versions, test inputs, sample size, evaluator configuration and prompt hashes.
- Inspect agent tool definitions, outbound allow rules, service endpoints, private links and network-denied traces.
- Inspect Foundry project and resource regions, model deployment regional availability, enterprise residency constraints and API error codes.
- Compare project connection configuration, model deployment resource ID, endpoint URL, deployment name and invocation authorization.
Explanation
Correct Answer
D:Compare project connection configuration, model deployment resource ID, endpoint URL, deployment name and invocation authorization. This is the most relevant evidence for foundry project links to wrong deployed model resource. A project workspace and a model deployment are distinct resources; using a logical model name without the correct deployment binding is insufficient. The matching controlled remediation is: Bind the application to the reviewed compatible model deployment in the correct Foundry project/resource and validate the connection.
Incorrect Answers
A:Compare experiment assignments, model deployment versions, test inputs, sample size, evaluator configuration and prompt hashes. This would investigate prompt a/b test uses different model versions in each arm rather than the documented foundry project links to wrong deployed model resource issue. The relevant investigation is: Compare project connection configuration, model deployment resource ID, endpoint URL, deployment name and invocation authorization.
B:Inspect agent tool definitions, outbound allow rules, service endpoints, private links and network-denied traces. This would investigate agent tool calls fail behind overbroad outbound block rather than the documented foundry project links to wrong deployed model resource issue. The relevant investigation is: Compare project connection configuration, model deployment resource ID, endpoint URL, deployment name and invocation authorization.
C:Inspect Foundry project and resource regions, model deployment regional availability, enterprise residency constraints and API error codes. This would investigate foundry project created in mismatched resource location rather than the documented foundry project links to wrong deployed model resource issue. The relevant investigation is: Compare project connection configuration, model deployment resource ID, endpoint URL, deployment name and invocation authorization.
Question 756
While deploying reproducible Azure Machine Learning infrastructure, the following condition occurs: An Azure ML pipeline consistently fails to read input blobs after a storage-account key rotation because its registered datastore uses the old key. The change must remain limited to authorized workspaces and avoid unreviewed production impact. What should the engineer examine first to confirm the specific AI-300 workload issue?
- Inspect compute cluster requested SKU, max nodes, available family quota, current utilization, node provisioning events and region availability.
- Review federated credential issuer, audience and subject, GitHub environment protection rules, branch patterns and run provenance.
- Inspect the datastore authentication mode, storage key rotation event, job input failure, and available identity-based storage permission assignments.
- Compare compute-instance schedules, last interactive activity, active notebook kernels, managed identity dependencies and cost reports.
Explanation
Correct Answer
C:Inspect the datastore authentication mode, storage key rotation event, job input failure, and available identity-based storage permission assignments. This is the most relevant evidence for datastore authentication relies on expiring account key. Datastores provide connection configuration but do not bypass storage authentication; credential rotation must be propagated or identity access designed correctly. The matching controlled remediation is: Replace the stale credential with a supported workspace identity-based datastore access path or rotate the authorized credential through secrets management.
Incorrect Answers
A:Inspect compute cluster requested SKU, max nodes, available family quota, current utilization, node provisioning events and region availability. This would investigate compute cluster cannot allocate requested gpu quota rather than the documented datastore authentication relies on expiring account key issue. The relevant investigation is: Inspect the datastore authentication mode, storage key rotation event, job input failure, and available identity-based storage permission assignments.
B:Review federated credential issuer, audience and subject, GitHub environment protection rules, branch patterns and run provenance. This would investigate oidc subject filter trusts unreviewed pull-request branches rather than the documented datastore authentication relies on expiring account key issue. The relevant investigation is: Inspect the datastore authentication mode, storage key rotation event, job input failure, and available identity-based storage permission assignments.
D:Compare compute-instance schedules, last interactive activity, active notebook kernels, managed identity dependencies and cost reports. This would investigate interactive compute instances remain running while unused rather than the documented datastore authentication relies on expiring account key issue. The relevant investigation is: Inspect the datastore authentication mode, storage key rotation event, job input failure, and available identity-based storage permission assignments.
Question 757
During a Foundry generative AI quality and observability incident, the following condition occurs: A Foundry agent returns an unhelpful answer after a search tool timeout but logs only the final generated text, hiding the actual failing service and request correlation. The fix must preserve evaluation validity and avoid unnecessary user-data collection. Which diagnostic evidence most directly distinguishes the described AI operations failure?
- Review user input field mapping, generated search query, response text, evaluator prompt and evaluation row samples.
- Examine trace field filters, application secret handling, workspace RBAC, log retention and representative captured payloads.
- Examine end-to-end request traces, retrieval timing, model time to first token, generation duration, request concurrency and percentiles.
- Inspect agent instrumentation, OpenTelemetry trace propagation, tool span durations, error attributes and correlated resource logs.
Explanation
Correct Answer
D:Inspect agent instrumentation, OpenTelemetry trace propagation, tool span durations, error attributes and correlated resource logs. This is the most relevant evidence for agent traces omit downstream tool failure reason. Multi-step AI agent troubleshooting needs per-tool traces, not just the final answer and an overall request status. The matching controlled remediation is: Enable scoped distributed tracing for agent and tool steps with error tags and privacy-safe identifiers. This matters here because a Foundry agent returns an unhelpful answer after a search tool timeout but logs only the final generated text,….
Incorrect Answers
A:Review user input field mapping, generated search query, response text, evaluator prompt and evaluation row samples. This would investigate relevance evaluator uses wrong intended user query rather than the documented agent traces omit downstream tool failure reason issue. The relevant investigation is: Inspect agent instrumentation, OpenTelemetry trace propagation, tool span durations, error attributes and correlated resource logs.
B:Examine trace field filters, application secret handling, workspace RBAC, log retention and representative captured payloads. This would investigate debug traces leak secrets and retrieved personal records rather than the documented agent traces omit downstream tool failure reason issue. The relevant investigation is: Inspect agent instrumentation, OpenTelemetry trace propagation, tool span durations, error attributes and correlated resource logs.
C:Examine end-to-end request traces, retrieval timing, model time to first token, generation duration, request concurrency and percentiles. This would investigate generation latency spike blamed on retrieval without breakdown rather than the documented agent traces omit downstream tool failure reason issue. The relevant investigation is: Inspect agent instrumentation, OpenTelemetry trace propagation, tool span durations, error attributes and correlated resource logs. This matters here because a Foundry agent returns an unhelpful answer after a search tool timeout but logs only the final generated text,….
Question 758
As an MLOps engineer reviews a model training or inference pipeline, the following condition occurs: A withdrawn credit-risk model is marked deprecated in a wiki but remains exposed as an active registry version and continues to be selected by automated deployment tools. The resolution must preserve quality gates, documented artifacts and rollback options. Which targeted Azure AI operations change best addresses the documented problem?
- Restrict access, redact or aggregate sensitive examples and publish a privacy-reviewed evaluation artifact for authorized reviewers.
- Archive or retire the version through a governed registry lifecycle and remove it from production deployment aliases after dependents migrate.
- Implement point-in-time correct feature retrieval and package its temporal lookup logic with the registered model version.
- Fix the image or artifact mount path and ensure init() loads the registered model before promoting the deployment.
Explanation
Correct Answer
B:Archive or retire the version through a governed registry lifecycle and remove it from production deployment aliases after dependents migrate. This targets retired model kept active as discoverable deployment candidate. Model lifecycle management must reflect state in the model registry and deployment references, not just informal documentation. Verify the failure through: Inspect registry model lifecycle status, alias assignments, endpoint references, deployment logs and retirement approvals. Validate success with: New deployment workflows cannot select the retired version while required historic audit artifacts remain available. This matters here because as an MLOps engineer reviews a model training or inference pipeline, the following condition occurs: A withdrawn….
Incorrect Answers
A:Restrict access, redact or aggregate sensitive examples and publish a privacy-reviewed evaluation artifact for authorized reviewers. This could address model evaluation exposes sensitive examples in explainability report, but it does not solve the specified retired model kept active as discoverable deployment candidate failure. The matching remediation is: Archive or retire the version through a governed registry lifecycle and remove it from production deployment aliases after dependents migrate.
C:Implement point-in-time correct feature retrieval and package its temporal lookup logic with the registered model version. This could address online feature retrieval pulls values from after prediction time, but it does not solve the specified retired model kept active as discoverable deployment candidate failure. The matching remediation is: Archive or retire the version through a governed registry lifecycle and remove it from production deployment aliases after dependents migrate.
D:Fix the image or artifact mount path and ensure init() loads the registered model before promoting the deployment. This could address endpoint deployment health probe fails before model initialization, but it does not solve the specified retired model kept active as discoverable deployment candidate failure. The matching remediation is: Archive or retire the version through a governed registry lifecycle and remove it from production deployment aliases after dependents migrate.
Question 759
During a Microsoft Foundry generative AI deployment review, the following condition occurs: A Foundry agent invokes its LLM successfully but receives access denied when querying an Azure AI Search index because the assigned managed identity lacks search data reader rights. The deployment must retain governed identities, endpoint contracts and auditable release history. Select TWO actions: one decisive diagnostic and its corresponding targeted remediation.
- Revoke the exposed key, use managed identity or approved secure Key Vault secret references, and scrub sensitive release configuration from ongoing builds.
- Compare candidate prompts using task-specific relevance and groundedness metrics alongside fluency before promotion.
- Inspect runtime managed identity, search service RBAC and index permissions, connector authentication, audit failures and client requests.
- Grant the narrowly scoped search data access role to the correct agent identity and use identity-based connection configuration.
- Examine model deployment quota for selected region/type, Azure CLI errors, requested capacity units and existing model allocations.
Explanation
Correct Answers
C:Inspect runtime managed identity, search service RBAC and index permissions, connector authentication, audit failures and client requests. This is the most relevant evidence for agent runtime managed identity cannot read retrieval data. Foundry model access and Azure AI Search data-plane authorization are separate permission boundaries for a retrieval-augmented agent. The matching controlled remediation is: Grant the narrowly scoped search data access role to the correct agent identity and use identity-based connection configuration. This matters here because a Foundry agent invokes its LLM successfully but receives access denied when querying an Azure AI Search index….
D:Grant the narrowly scoped search data access role to the correct agent identity and use identity-based connection configuration. This targets agent runtime managed identity cannot read retrieval data. Foundry model access and Azure AI Search data-plane authorization are separate permission boundaries for a retrieval-augmented agent. Verify the failure through: Inspect runtime managed identity, search service RBAC and index permissions, connector authentication, audit failures and client requests. Validate success with: Authorized agent queries retrieve permitted indexed chunks while direct access to unrelated indexes remains denied. This matters here because a Foundry agent invokes its LLM successfully but receives access denied when querying an Azure AI Search index….
Incorrect Answers
A:Revoke the exposed key, use managed identity or approved secure Key Vault secret references, and scrub sensitive release configuration from ongoing builds. This is associated with foundry infrastructure parameters embed a secret in source repository, not the essential evidence-and-fix pair for agent runtime managed identity cannot read retrieval data. First examine: Inspect runtime managed identity, search service RBAC and index permissions, connector authentication, audit failures and client requests. Then correct: Grant the narrowly scoped search data access role to the correct agent identity and use identity-based connection configuration.
B:Compare candidate prompts using task-specific relevance and groundedness metrics alongside fluency before promotion. This is associated with prompt variant comparison measures fluency not task success, not the essential evidence-and-fix pair for agent runtime managed identity cannot read retrieval data. First examine: Inspect runtime managed identity, search service RBAC and index permissions, connector authentication, audit failures and client requests. Then correct: Grant the narrowly scoped search data access role to the correct agent identity and use identity-based connection configuration.
E:Examine model deployment quota for selected region/type, Azure CLI errors, requested capacity units and existing model allocations. This is associated with azure cli foundry setup assumes existing model quota, not the essential evidence-and-fix pair for agent runtime managed identity cannot read retrieval data. First examine: Inspect runtime managed identity, search service RBAC and index permissions, connector authentication, audit failures and client requests. Then correct: Grant the narrowly scoped search data access role to the correct agent identity and use identity-based connection configuration.
Question 760
During a Microsoft Foundry generative AI deployment review, the following condition occurs: A new troubleshooting prompt sounds polished in sample responses but fails to give actionable next steps because the evaluation measures only fluency and omits task completion and grounding. The deployment must retain governed identities, endpoint contracts and auditable release history. Which acceptance test shows the AI operations defect is corrected?
- The selected prompt improves successful issue resolution without factual regressions on the held-out support dataset.
- Adversarial test documents no longer cause unauthorized agent actions while the assistant still answers legitimate questions from the content.
- Requests from the approved VNet resolve the private address and succeed, while public connectivity remains blocked.
- Every production prompt version resolves to a reviewed commit and a failed change can be rolled back without copying an unknown portal draft.
Explanation
Correct Answer
A:The selected prompt improves successful issue resolution without factual regressions on the held-out support dataset. A fluent response can still be wrong or unhelpful; prompt tuning requires evaluation metrics tied to application goals. a direct corrective action is to compare candidate prompts using task-specific relevance and groundedness metrics alongside fluency before promotion. This matters here because a new troubleshooting prompt sounds polished in sample responses but fails to give actionable next steps because the….
Incorrect Answers
B:Adversarial test documents no longer cause unauthorized agent actions while the assistant still answers legitimate questions from the content. That addresses a different configuration issue rather than the main requirement. Verify instead: The selected prompt improves successful issue resolution without factual regressions on the held-out support dataset.
C:Requests from the approved VNet resolve the private address and succeed, while public connectivity remains blocked. That addresses a different configuration issue rather than the main requirement. Verify instead: The selected prompt improves successful issue resolution without factual regressions on the held-out support dataset.
D:Every production prompt version resolves to a reviewed commit and a failed change can be rolled back without copying an unknown portal draft. That addresses a different configuration issue rather than the main requirement. Verify instead: The selected prompt improves successful issue resolution without factual regressions on the held-out support dataset. This matters here because a new troubleshooting prompt sounds polished in sample responses but fails to give actionable next steps because the….